Depth sensing can be accomplished through active or passive depth sensing systems. In active sensing, a known pattern can be projected into a target scene, often referred to as “structured light.” The structure of the pattern projected into the target scene encodes depth information for objects in the target scene. To identify depth information, the pattern can be found in a captured image of the target scene and a three-dimensional representation of the target scene can be reconstructed based on the relationship between the known projected pattern and the decoded pattern. Examples of passive depth sensors include stereoscopic imaging systems. In stereoscopic imaging, multiple images of the same target scene are captured from slightly different angles or perspectives. Stereo matching techniques can use disparity between the images to generate a depth map of the target scene.